While we know much about the neuronal mechanisms that underlie simple reward-directed behaviors, such as conditioning, we know very little about how reward information interacts with executive processes such as working memory. The aim of this grant is to examine how different rewards are held in working memory, compared and then selected, in order to allow efficient, flexible behavior. Monkeys will be trained to choose one of two sequentially presented rewards, separated by a brief delay. Thus, the monkey has to hold the first reward in working memory until the second reward is presented, at which point it compares the two rewards and selects one for its choice. We will record the activity of single neurons in the prefrontal cortex (RFC) while the monkey is performing this task. We will examine whether RFC neurons maintain information about the reward during the delay (indicative of working memory processes), and whether this encoding is relatively abstract and flexible which would facilitate the evaluation of widely disparate rewards and goals. We will investigate the neuronal processes that underlie the comparison and selection of the two rewards, processes which are important components of decision-making. We will also compare how primary and secondary reinforcers are held in working memory. Dysfunction of the RFC, particularly the orbitofrontal region, has been implicated in drug addiction. A failure to use working memory representations of alternate rewards and goals (such as quitting the drug) might exacerbate addiction by preventing the addict from exerting top-down control over addictive behavior. Thus, understanding the neuronal mechanisms that determine how choices and priorities are made between different rewards may provide us with a clearer understanding of the pathology of addiction, and with alternative therapeutic strategies aimed at boosting top-down control and attenuating the likelihood of relapse in the recovering addict.